Kata

Yeah I saw that after I posted the issue, but decided to leave it up because although this problem elects to ignore that it's called "matrix" rotation. To my understanding in programming the term matrix is not applied to one dimensional arrays, and outside of things like C# multi-dimensional arrays the only thing that would actually meet that criteria is nested arrays.

"the only thing that would actually meet that criteria is nested arrays." => Well, no. For instance, an array of string does the job just right. One could also choose to represent a whole 2D matrix with just one string (i.e https://www.codewars.com/kata/586214e1ef065414220000a8 and other string input katas do exactly that). I'm not sure what makes you think something would be wrong.

Thank you! :D

Argh! Lost my answer in the limbs of Internet! Well, thanks for making me realize that I did it too quickly and didn't take enough time to make it clearer. So, what did go wrong?

You are not supposed to modify the input array/list.
It's just that I (originally) used the same matrix as input for your solution and the validator, and when blind4Basics told me that this exposes the kata to hacks I replaced it with the test that you failed.

Which language was it though? It might not be very clear either but it should have shown you the expected value of the parameter array (unchanged) and the array after calling your function.

Edit: Well, I can't update the kata right now because the server times out while trying o_O

Kata edited! The description and tests ouput should be clearer as what's going wrong. Also added input safety check directly in the example testcases so you can see what's going on.

I'm in Javascript, and it showed the expected vs my outcome. Usually, though, I can put a console.log(matrix) line in the beginning of my code to see what the input is for the random tests (which could make me see what the connection is between these failed cases) but for some reason that's not showing up... maybe a side consequence of avoiding the cheating?

I'll have a crack with the new description and test cases though, thanks for taking a look! Also really enjoyed the Kata, sorry for just whinging at you here in the comments...

I see, managed to fix it! Thanks. My solution did modify the input (splitting the elements into arrays), so I just added a 'for' loop to the beginning to copy the input to a new array and the code worked for all.

In terms of wording your warning in the description, I'd suggest writing something like 'if your code modifies 'matrix' then the tests won't work correctly, if you need to do this you must create a copy of the input to keep 'matrix' intact at the end of the solution' or something along those lines?

Purely for satisfaction level during the kata, it would also be cool to have more test cases like the final one to see some lovely rotation in action! Thanks for the kata.

Ah yes of course you can't print before these specific tests because they're appendices to the tst right before them. Dunno if I should wrap the couple in a it though, it could make the ouput slightly overcrowded.
Edit: Nah it's just clearer to wrap every parts of a single test in a it. Description updated and added a few other arrays as well.

Should be corrected, I thought the whole validation part was wrapped and only the preloaded code was accessible :blush:

Edit : I don't know why the cheating solutions aren't marked as invalid, even though I checked twice that they won't solve the kata anymore o_O

Edit 2: I strongly disagree with this statement:

Last but not least, think about the problems that could appear when a warrior is mutating your input during the random tests (if you pass lists, dictionaries,...). Two ways to avoid problems:

• compute the expected result first, before passing the input to the warrior
• or pass a copy of the input to the warrior (use copy.deepcopy if needed)

I personally think it's very important that the user does NOT alter the input, unless explicitely stated otherwise. And it thus SHOULD return an error if that's the case (though I'll admit that the error log will be somewhat confusing).

And btw there's a typo in this section:

from copy import deepcopy
test.assert_equals(deepcopy(user_func(input)), ref_func(input))

should be

from copy import deepcopy
test.assert_equals(user_func(deepcopy(input)), ref_func(input))

Otherwise the copy is kinda useless ;)

Ok you're right I didn't think this through. Corrected that part as well.

yeah, at first I thought it could be interesting to let the matrixGenerator and array testing function for user's test cases. But it's actually not that interesting and I even forgot to mention it in the description so I should just put it in the test case block.

Thanks! my answer seems to have disappeared so I'll say it again, I was looking for that kind of function but I never know which functions are implemented from the test libraires in codewars. Didn't know what to output when using 'expect' so I tried almost everything I coudl think of ;P
Thank you for your thorough testing as well, I'm a little too reckless sometimes ^^'